An amplifier, particularly an integrated amplifier

Description

Field of the invention

[0001] The present invention relates to amplifiers and has been developed with particular attention to its possible use in monolithic integrated amplifiers, the inputs of which are required to accept common-mode voltages which go below the earth voltage but which, at the same time, have high input common-mode rejection ratios (CMRR), power-supply rejection ratios (PSRR) and input impedance. The need for such an amplifier arises, for example, in the production of circuits for interfacing between the zirconium dioxide probes (lambda probes) fitted to catalytic converters for motor vehicles with internal combustion engines and the central electronic control units which are fitted in the motor vehicles for controlling their engines.

[0002] In most applications, the signal of the lambda probe relates to the earth of the converter and varies from 0 to a maximum of about 1 volt. The earth of the probe (which corresponds to that of the converter) is generally different from that of the engine control unit in which the amplifier is housed; this difference is within ± 1 volt. Since the connecting wires between the probe and the control unit are long, it is necessary to have a high CMRR in order to reject common-mode interference. Finally, the amplifier has to output a signal which relates to earth and which has an optimal dynamic range for an analog/digital converter (0-4 volts), which means it must have a nominal gain of 4. The probe has a high output impedance and the input impedance of the amplifier therefore has to be high.

Description of the prior art

[0003] A class of amplifiers which is particularly suitable for satisfying the aforementioned characteristics is that of the so-called "Instrumentation Amplifiers".

[0004] In this connection, the following references may usefully be consulted:

[0005] Rudy J. Van de Plassche, "A wide-band monolithic instrumentation amplifier", IEEE J. Solid Circuits, Vol. SC-10, pp. 424-431, Dec. 1975, A. Paul Brokaw, Michael P. Timko, "An improved monolithic instrumentation amplifier", IEEE J. Solid Circuits, Vol. SC-10, pp. 417-423, Dec. 1975, and Bernard D. Miller, Lawrence R. Sample, "Instrumentation amplifier IC designed for oxygen sensor interface requirements", IEEE J. Solid State Circuits, Vol. SC-10, Dec. 1981.

[0006] Also, in EP-A-0 045 841 a voltage-current converter is disclosed having an input related to ground and an output current which is taken up by the device. The linear proportionality constant is obtained by using exclusively MOS transistors, which are far from being inherently linear.

[0007] Furthermore, a wide-band current amplifier providing very low input resistance, good large signal behaviour and large bandwidth is disclosed in an article by Z. Wang entitled "Wideband Class AB (Push-Pull) Current Amplifier in CMOS Technology" which appeared in Electronics Letters, Vol. 26, No. 8, 14 April 1990, ENAGE GB, pages 543-545. In that reference, both the input signal and the output signal are current signals. The power supply rejection is low, similar to that of a traditional current-mirror: to improve this a cascode is added, as shown in Figure 2 of the reference.

[0008] Most amplifiers of the type specified above and, in particular, those described in the first two references cited above, generally have the structure illustrated in Figure 1, with a first stage formed by two transistors T1 and T2, the emitters of which are coupled by means of a resistor RG and are connected to respective current generators I1 and I2, and the collectors of which are associated with respective resistors R'1 and R'2. Similarly, the emitters of transistors T3 and T4 of the second stage are coupled by a resistor RS and are connected to respective current generators I3 and I4, and respective resistors R'3 and R'4 are associated with their collectors.

[0009] The input signal VIN is applied between the bases of the transistors T1 and T2 and a differential amplifier A takes a signal (an output signal Vout) from between the collectors of T1 and T2 and transfers this signal also to the base of T4. A further amplifier B transfers the signals present at the collectors of T3 and T4 as driving signals to current generators I2, I4 and I1, I3.

[0010] These amplifiers may have high input impedance as well as high CMRRs and PSRRs. They cannot be used in the context indicated above, however, since the minimum value of the common-mode voltage of the inputs is equal to Vbe+Vce(sat)≃+1V relative to earth (as against the -1V required).

[0011] The solution presented in the article by Miller and Sample (which is also documented in United States patent No. 4,232,271) differs substantially from the other known solutions precisely in order to satisfy the input common-mode dynamic-range requirement. This amplifier is dedicated for use, preferably, with a lambda probe. The input stage (in this connection, see the detailed description provided in the United States patent No. 4,232,271, Figure 4 of which is reproduced as Figure 2), is formed by the emitters of transistors 18 and 19 which may reach a voltage below that of the earth, since the voltages of their collectors are kept above the earth potential by means of amplifiers 31, 32 and the reference voltage VREF, thus preventing the substrate diode from conducting (it should be remembered that the amplifiers in question are intended to be integrated). The high input impedance is achieved by biasing the transistors 18 and 19 with a first generator formed by the reference voltage VREF, the resistor 21 and the amplifier 31 and a second generator formed by the same voltage VREF, the resistor 20 and the amplifier 32, respectively. The voltage applied to the inputs (VIN) is transformed into a current I1 = VIN/R23, where R23 indicates the resistance of the resistor 23. This current is then mirrored, by means of the circuit formed by the transistors 29b and 33 and by the resistors 35 and 36, into the resistor 16. With this structure, it is necessary, in order to achieve high CMRR and PSRR values, to use a "cascode" for the current-mirror circuit cited above, as well as for the generators of the currents I1 = VREF/R21 and I2 = VREF/R20 (R21 and R20 are the resistances of the resistors 20 and 21). In fact, the voltages at the terminals of these current generators vary with the input voltage (both common-mode and differential). Although, a "cascode" is possible for I1 and I2, even with a single 5-volt supply, it is not possible for the current-mirror circuit including the transistors 29b and 33 and the resistors 35 and 36 for which, with the use of mirror circuits known in the art, it is necessary to use a second supply in order to have a "cascode" and at the same time not to have errors due to the current for biasing the transistors which are connected to the collector voltage Vcc, and to permit a high input common-mode dynamic range.

Objects and summary of the invention

[0012] The primary object of the invention as it is defined in the appended claims is thus to avoid the use of a second supply (for example, in the case of installation in a motor vehicle, the battery) so that the amplifier can be supplied with a single 5-volt supply, naturally whilst retaining the characteristics of high input impedance and dynamic range with equally high CMRR and PSRR values.

[0013] More specifically, according to the present invention, this object is achieved by virtue of an amplifier having the characteristics recited in the following claims.

Detailed description of the invention

[0014] The invention will now be described, purely by way of non-limiting example, with reference to the appended drawings, in which:

Figures 1 and 2, which relate to the prior art, have already been described above,

Figure 3 shows the structure of the circuit of an amplifier according to the invention, and

Figures 4 to 7 show in detail the context of application and some structural details of the invention.

[0015] The general circuit diagram of the amplifier according to the invention is indicated 1 in Figure 3.

[0016] Within this structure, the input signal is applied between two terminals indicated IN(-) and IN(+) respectively, which are connected to the emitters of two bipolar transistors Q1 and Q2, the bases of which are coupled by means of a resistor R1. The base of the transistor Q1 is connected, on the one hand, to the emitter of the same transistor, by means of a resistor R2 and, on the other hand, to the output of a differential amplifier A1, the inverting input of which is connected to a reference voltage VR1 and the non-inverting input of which is connected to the collector of the transistor Q1. The same collector is also supplied by a current generator IP3 which is connected to the general supply line of the amplifier, which is assumed to be brought to a supply voltage Vcc, for example, of the order of 5 volts, which is common to all the elements of the circuit. A substantially identical connection arrangement is adopted for a current generator IP4 which supplies the collector of the transistor Q2. The same collector is connected to the inverting input of a further differential amplifier A2, the non-inverting input of which is connected to a further reference voltage VR2. The output of the amplifier A2 is connected to the gates of three MOS transistors M7, M10 and M9. The drain electrode of the MOS M7 transfers a current IR1 to the base of the transistor Q2 and hence to the first resistor R1 and the source electrode of the same transistor M7 is connected to the collector of a further bipolar transistor Q5, the emitter of which is connected, by means of a resistor R3, to the common supply line Vcc of the circuit.

[0017] The drain electrode of the MOS M10 is connected to the earth G of the circuit and the source of the same MOS M10 is connected to the collector of a further bipolar transistor Q6, the emitter of which is connected to the common supply line Vcc of the circuit, by means of a resistor R4. It will be noted that the collector and the base of the bipolar transistor Q6 are short-circuited.

[0018] Finally, the drain of the MOS M9 is connected to the earth G by means of a resistor R6, from the terminals of which the output voltage VOUT of the amplifier is taken. The source of the MOS M9 is connected to the collector of a bipolar transistor Q7, the emitter of which is connected to the supply line Vcc by means of a resistor R5.

[0019] In the circuit configuration described, the emitter of Q1 can go below the earth voltage since its collector is kept at the voltage VR1 by means of A1. Similarly, the collector voltage of Q2 is kept at VR2 by means of A2 and the MOS M7.

[0020] Since the bipolar transistors Q1 and Q2 are biased by two constant-current generators (IP3 and IP4), their base-emitter voltages (Vbe) may be considered approximately constant. In these circumstances, an input signal VIN applied as the voltage difference between IN(+) and IN(-) is transferred to the terminals of R1. In practice, the voltage, indicated VR1, at the terminals of R1 is approximately equal to the input voltage VIN.

[0021] The current necessary for the voltage VR1 to be present at the terminals of R1 is supplied by the MOS M7 and is equal to IR1 = VR1/R1, in which, naturally, R1 is the resistance of the corresponding resistor.

[0022] The current IR1 is supplied by M7 through the current-mirror circuit formed by the bipolar transistors Q5, Q6, Q7 and the MOS transistors M7, M9 and M10. The gate voltage of M10 is regulated by the output of A2 so that, after it has been mirrored in the branch formed by the transistor Q5, the MOS M7 and the resistor R3, the current in the branch including the MOS M10 and the bipolar transistor Q6 and the resistor R4 equals the current IR1 calculated above (taking account of the ratio between the areas of Q5 and Q6 and of the ratio between the values of R3 and R4). In these circumstances, if the area of Q7 is made equal to that of Q5 and the value of R5 is made equal to that of R3, the current in Q7 will be equal to IR1 without any error due to the bias currents of Q5, Q6 and Q7. IR1 is thus mirrored into R6 without error due to the bias current and with a "cascode" both at the input and at the output of the mirror formed by the MOS transistors, which make its resistance greater than would be the case if bipolar transistors were used.

[0023] If the ratio between the currents in the mirror is 1:1, the current in R6 (IR6) will be equal to that in R1, so that:
IR1 = IR6
VOUT = R6.IR1
VOUT = R6/R1.VIN

[0024] In other words, according to the terminology adopted in the claims which follow, IR6 constitutes a "replica" (identical, as in the example illustrated, or possibly amplified or attenuated) of IR1.

[0025] The maximum dynamic range at the input IN(+) is limited by the minimum voltage required in order for the input of the circuit for mirroring the current IR1 into R6 to be in its linear region. This minimum voltage is:

in which:

VSDM7min represents the minimum value of the voltage between the source and the drain of the MOS M7,

VECQ5 represents the voltage between the emitter and the collector of the transistor Q5, and

VR3 represents the voltage at the terminals of the resistor R3.

[0026] Assuming, for convenience, that all the resistors and the MOS and bipolar transistors in question are the same, it may be considered that:

• VECQ5 is approximately equal to VEBQ6 (the base-emitter voltage of the transistor Q6) and hence approximately 0.7 volts,
• a voltage drop in R3 of about 200 millivolts is possible, since the maximum current in R3 can easily be deduced by the foregoing equations, from the maximum input voltage to be admitted, and
• a VDSM7 SAT (the voltage between the drain and the source of the MOS M7 in the saturation condition) of about 200 millivolts is possible if a suitable size for M7 is selected, in dependence on the maximum current IR1, which may be considered approximately equal to the maximum current R3 calculated in the previous point.

[0027] Under these circumstances, the maximum voltage, indicated VIN(+)max, which can be applied to the terminal IN(+), will be given by the equation:

in which Vcc is the common supply voltage of the circuit and VBEQ2 represents the base-emitter voltage of the transistor Q2. Assuming that Vcc is 5 volts and VBEQ2 is about 0.7 volts, VIN(+)max is thus about 3.2 volts.

[0028] This value is suitable for a lambda probe 10 fitted to a catalytic converter K (see in particular Figure 4) and intended to be connected to a central control unit, the amplifier 1 acting as the input stage of the control unit and the output signal VS of the probe being taken from between its "hot" electrode and the earth G' of the converter K for application between the terminals IN(+) and IN(-), with the latter terminal at a voltage VM(±1 volt) relative to the earth G.

[0029] In particular, it is assumed that the lambda probe 10 brings to the input IN(-) a maximum voltage VMmax given by:

with a maximum value of the signal Vs, indicated VSmax, of:

which gives as a result:

[0030] Thus, it is possible to have a dynamic range suitable for the lambda probe, without using a supplementary voltage, whilst making use of a "cascode" mirror circuit. The restrictions on the input dynamic range IN(+) enable the calculation of minimum value VCQ2min of the dynamic range at the collector of Q2, which is:

in which VCEQ2sat indicates the saturation collector-emitter voltage of Q2 (about 0.2 volts).

[0031] This means that VR2 can be fixed at a value just greater than VCQ2min, and the current generators IP3 and IP4 can thus have wide dynamic ranges. The maximum dynamic range in R6 is limited by the dynamic range of the branch constituted by M9, Q7 and R5. This voltage is determined by the minimum voltage required at the output of the circuit for mirroring the current IR1 into R6 which, if the current ratio IR1/IR6 is unitary, corresponds to VSm, indicated above. This gives a maximum output voltage VOUTmax of:

[0032] With this dynamic range, a gain of 4 cannot be achieved in R6; if VSmax = 1V, VOUTmax would in fact have to be at least 4V. It is therefore necessary to place in cascade an amplifier stage which outputs the required gain and dynamic range.

[0033] The reference voltages VR1 and VR2 are more effective, as regards the PSRR, if they are independent of the supply. This can be achieved by means of a structure such as that shown in Figure 5, which may be considered known. In this drawing, VBG indicates a reference voltage independent of the supply voltage Vcc. The voltage VBG may be a "bandgap" voltage or the voltage drop in a resistance of which one end is connected to earth and the other end is supplied by a current generator independent of the supply. For this purpose, the reference voltage VRi (where i = 1 or 2 in the diagram of Figure 3) is thus taken from the output of a differential amplifier A10, to the non-inverting input of which VBG is applied, a resistor R10 being connected between the output and the inverting input of the same amplifier A10 and another resistor R20 being interposed between the inverting input and the earth G of the circuit. The amplifier A10 is supplied by the general voltage supply Vcc.

[0034] Finally, in order to have a high CMRR value and a low input offset voltage, it is necessary to prevent the absorption of current by the collectors of Q1 and Q2, which are already biased at very low current values (typically 0.5 µA), so as to have a high direct input resistance and thus to avoid charging of the probe. For this purpose, it is preferable to form the amplifier stages A1 and A2 of Figure 3 with the use of the circuits shown in greater detail in Figures 6 and 7, in which the structures of the amplifiers A1 and A2 in question are shown with reference to their connections within the circuit of Figure 3.

[0035] In particular, both the amplifiers in question are formed with MOS transistor input stages, indicated M8 for the amplifier A1 of Figure 6, and M5 for the amplifier A2 of Figure 7, respectively.

[0036] With reference, in particular, to the amplifier A1, it will be noted that the gate of the MOS M8 is intended to be connected to the collector of the transistor Q1 and the drain of the MOS M8 is connected to the supply line Vcc.

[0037] The source of the MOS M8, however, is connected to the emitter of a further bipolar transistor Q9, the collector of which is connected to the base of the transistor Q1 and hence to the opposite lead of the resistor R1 from the transistor Q2.

[0038] The maximum voltage at the input IN(-) which the amplifier makes available to the probe (which, as has been seen, would require only 1 V), can be expressed by the equation:

in which, naturally, VBEQ1 represents the base-emitter voltage of the transistor Q1, VECQ9SAT represents the voltage between the emitter and the collector of the transistor Q9 in the saturated condition, and VEBQ9 represents the emitter-base voltage of the same transistor.

[0039] If it is assumed that VR1 is 2.5 V and VBEQ1 is approximately equal to VEBQ9, with VECQ9SAT about 0.3 V, a VIN(-)max of about 2.2 V is obtained.

[0040] The maximum fixed voltage VCQ1max at the collector of Q1, however is:

in which, naturally, VGSM8 represents the voltage between the gate and the source of the MOS M8.

[0041] This value does not permit a "cascode" to be formed in IP3 since if Vcc = 5V, the dynamic range available would be too small. In practice, the voltage VIP3_max at the terminals of P3 would have a value which can be expressed by the equation:

in which VCQ1max represents the maximum value of the voltage at the collector of Q1.

[0042] If A1 is formed as shown in Figure 6, in order to have a high output impedance, IP3 must therefore be formed with a transistor, the emitter of which is degenerated by a resistance. If it is essential to form the generators IP3 and IP4 with "cascodes", it is necessary to change the structure of A1 so as to reduce the voltage difference between its non-inverting and inverting inputs. Naturally, this can be done by increasing the gain of A1 which will therefore require a complex structure needing a larger area of silicon.

[0043] As regards the structure of A2 (Figure 7), it will be noted that the drain of the MOS M5 is connected to the collector of a bipolar transistor Q3, the base and the collector of which are short-circuited and the emitter of which is connected to the earth G. The source of the MOS MS is supplied by a current generator IP19 which is connected to the supply voltage Vcc which also supplies the source of another MOS M6, the gate of which is connected to the reference voltage VR2. The drain of the MOS M6 is connected to the collector of a bipolar transistor Q4, the base of which is connected to the base of the transistor Q3 and the emitter of which is connected to the earth G. The collector of the transistor Q4 (and hence the drain of the MOS M6) drives the base of a further bipolar transistor Q11, the emitter of which is connected to the earth G and the collector of which is intended to drive the gate of the MOS M7 (and thus constitutes the output of the amplifier A2) and is supplied by a further current generator IP43 which is connected to the supply voltage Vcc.

[0044] As stated, the advantages of the invention consist essentially of the fact that it provides an amplifier, particularly an amplifier which can be integrated, which has high CMRR and PSRR values, with inputs which can move relative to the earth by plus or minus 1V in common mode and which, at the same time, can accept a maximum difference of 1V between the input IN(+) and the input IN(-), the amplifier having a single supply Vcc of the order of 5V and a high input impedance.

[0045] This is all made possible by a set of circuitry measures which can be summarised in the following terms.

[0046] In the first place, bipolar input transistors Q1 and Q2 are used and are biased with constant currents, their emitters being connected to the inputs IN(+) and IN(-) which can fall to a voltage below that of the earth, since their respective collectors are kept at fixed reference voltages VR1 and VR2 (which are positive relative to the earth) by the amplifiers A1, A2, thus preventing the substrate diode from conducting.

[0047] Moreover, the selection of a very low value (less than 1 µA, for example 0.5 µA) for the currents produced by the generators IP3 and IP4 which supply the collectors of the input transistors Q1 and Q2, enables the direct current to be absorbed by the input generator at the input IN(+), to be only 0.5 µA, whilst, for alternating current, the high resistances of the inputs are limited essentially by the so-called R_O (or 1/h_oe) of the transistors Q1, Q2 and by the gains of the amplifiers A1 and A2.

[0048] Thirdly, a circuit is used to reflect the current IR1 from the resistance R1 into the resistance R6, which has a terminal connected to the earth. This is achieved by controlling the gate voltage of the MOS M10. A variation in this voltage will vary the voltage between the gate and the source of M10, thus varying the current in the transistor Q6 and the current in the resistor R4, which constitute the input of a mirror. A first output of the mirror, which is constituted by the MOS M7, the transistor Q5 and the resistor R3 (the first branch of the circuit), will supply the current IR1 which is controlled by the feedback loop by means of the gain of the amplifier A2, in dependence on the input voltage in the terms described in detail above. A second output of the mirror is constituted by the MOS M9, the transistor Q7 and the resistor R5 (the second branch of the circuit) in order to supply current to the resistor R6. The ratio between the areas of Q5 and Q7 and the ratio of the resistances of R3 and R5 determines the ratio between the current in R6 and the current IR1 so that either a unitary value or a value other than 1 may be attributed to this ratio.

[0049] At the same time, the two branches mentioned are nor affected by errors due to the currents for biasing the bases of the transistors Q5 and Q7, since both collect at the collector of Q6 (which, together with the MOS M10 and the resistor R4, define a third arm of the current-mirror circuit, which is structurally identical to the first two branches). In particular, although the current in Q6 is the input current of the mirror, it may differ both from IR1 and from the current in R6. The current in R6 is thus the current IR1 multiplied by the ratio indicated above, without errors due to the bias currents, the error due to the difference between the base-emitter voltages of the transistors Q5 and Q7 being attenuated by means of the resistors R3 and R5 at their respective emitters. Moreover, the current in R6 has an error due to the Early voltages of the transistors Q5 and Q7 and this is attenuated by the "cascode" formed by the MOS M7 and the MOS M9. This is achieved whilst retaining a high input dynamic range at the drain of the MOS M7. The use of an MOS transistor (or any other similar device with a negligible input bias current) for M7 increases the common-mode rejection rates of the inputs.

[0050] Moreover, an amplifier A1 with high input impedance is used and, with the dynamic-range characteristics given by a sensor such as a lambda probe, and with a supply voltage of the order of 5V, can be optimised as shown in Figure 6. In this diagram, the MOS M8 prevents the absorption of current by the collector of the transistor Q1. No offset voltage is thus introduced to the inputs IN(+) and IN(-). Moreover, since M8 is a MOS transistor, the CMRR value is increased; in fact, a variation in the current in Q9 and M8 as a result of the Early effect due to Q9 causes no variation in the bias current of Q1 (which is fixed by IP3) and no variation in the current in the mirror and hence at the output.

[0051] Moreover, the use, as the amplifier A2, of a structure with high input impedance formed with MOS transistors (M5, M6 in Figure 7) with a zero input bias current avoids the introduction of an offset voltage at the inputs IN(+) and IN(-).

[0052] Naturally, the principle of the invention remaining the same, the details of construction and forms of embodiment may be varied widely with respect to those described and illustrated, without thereby departing from the scope of the present invention. In particular, although the selection of components such as bipolar transistors or MOS transistors in certain positions in the circuit is considered greatly preferable for the purposes of putting the invention into practice, it should not be considered absolutely binding for the purposes of the implementation thereof. In particular, the replacement of one or more of these components by functionally equivalent components is certainly possible and is therefore included within the scope of the present invention.

[0053] This applies particularly of the possibility, for example, of replacing bipolar transistors with field-effect transistors (for example MOSFETs) and vice versa. In this connection, for the purposes of defining the invention, the terms "collector", "base" and "emitter" as used in the following claims should be considered equivalent to the terms "drain", "gate" and "source", respectively, as far as the possibility of using one type of transistor instead of another is concerned. Naturally, the currently-preferred embodiment provides specifically for some functions to be entrusted to bipolar transistors and others to MOS transistors. In any case, as already stated, although this selection is preferred, it should not be considered essential.

Claims

1. An amplifier, particularly an integrated amplifier, including, in combination:

- an input stage including two transistors (Q1, Q2) biased by constant currents (IP3, IP4), with their respective collectors kept at constant, generally positive, fixed voltage levels (VR1, VR2) so that their respective emitters, which define the input terminals (IN(+), IN(-)) of the amplifier, can fall to a voltage below that of the earth, preventing the substrate diode from conducting, the two transistors (Q1, Q2) being connected by their respective bases by means of a first resistor (R1) so that, in use, a voltage corresponding to the signal (VIN) applied between the input terminals (IN(+), IN(-)) is applied to the terminals of the first resistor (R1),

- a current-mirror circuit (Q5, Q6, Q7, M7, M9 and M10) including first means (M7) for generating, in the first resistor (R1) a first current (IR1) corresponding to the voltage applied to the terminals of the first resistor (R1) and second current-generating means (M9) which are coupled to the first current-generating means (M7) in a mirror relationship in order to supply at least one second resistor (R6) with a second current (IR6) which constitutes a replica of the first current (IR1), so that an output signal (VOUT) corresponding to the input signal (VIN) can be taken from the terminals of the second resistor (R6), and

- a common supply line (Vcc) for the input stage (Q1, Q2) and for the current-mirror circuit (Q5, Q6, Q7, M7, M9 and M10), characterised in that it includes

   amplifier means (A2) which controls the first current-generating means (M7) in dependence of the input signal (VIN), and in that
   said current mirror circuit includes first, second, and third branches, wherein :

- the first branch (M7, Q5, R3) defines the first current-generating means (M7) and is the output of the current mirror generating said first currrent (IR1) in said first resistor (R1),

- the second branch (M9, Q7, R5) which defines the second current-generating means (M9) and is coupled to the first branch (M7, Q5, R3) in a current-mirror relationship in a generally "cascode"-like configuration, and

- the third branch (M10, Q6, R4) is structurally symmetrical to the first branch (M7, Q5, R3) and to the second branch (M9, Q7, R5), the third branch being interposed between the common supply line (Vcc) and the earth (G) of the circuit, the first branch (M7, Q5, R3) being interposed between the common supply line (Vcc) and the first resistor (R1), and the second branch (M9, Q7, R5) being interposed between the common supply line (Vcc) and the second resistor (R6), and further wherein :

each branch of the current-mirror circuit includes :

- a first transistor (Q5, Q6,Q7) with its emitter connected to the common supply line (Vcc) and its base connected to the bases of the homologous transistors of the other branches of the current-mirror circuit, and

- a second transistor (M7, M10, M9), preferably constituted by a MOS transistor, with its source connected to the first transistor (Q5, Q6, Q7) and its gate connected to the gates of the homologous transistors of the branches of the current-mirror circuit,

the base of the first transistor (Q6) included in the third branch (M10, Q6, R4) being short-circuited to one of the other terminals of the first transistor (Q6) in a generally diode-like configuration.

2. An amplifier according to Claim 1, characterised in that one terminal of the second resistor (R6) is connected to earth.

3. An amplifier according to Claim 1 or Claim 2, characterised in that the two transistors of the input stage (Q1, Q2) are associated with constant-current biasing circuits (IP3, IP4) which supply constant bias currents in the µA range.

4. An amplifier according to Claim 3, characterised in that the two transistors of the input stage (Q1, Q2) are associated with constant-current biasing circuits (IP3, IP4) which supply a constant bias current of less than 1 µA, preferably of the order of 0.5 µA.

5. An amplifier according to any one of the preceding claims, characterised in that the two transistors in the input stage (Q1, Q2) are bipolar transistors.

6. An amplifier according to Claim 1, characterised in that the drains of the second transistors (M7, M9) included in the first (M7, Q5, R3) and second (M9, Q7, R5) branches of the current-mirror circuit are connected to the first resistor (R1) and to the second resistor (R6), respectively.

7. An amplifier according to Claim 1, characterised in that the drain of the second transistor (M10) included in the third branch (M10, Q6, R4) of the current-mirror circuit is connected to the earth (G) of the current-mirror circuit.

8. An amplifier according to Claim 1, characterised in that the first transistor (Q5, Q6, Q7) is a bipolar transistor.

9. An amplifier according to Claim 1, characterised in that the first current-generator means are constituted substantially by a MOS transistor (M7) or any other similar device with a negligible input bias current, so as to increase the common-mode rejection ratio of the inputs of the amplifier (IN (-), IN (+)).

10. An amplifier according to Claim 1, characterised in that the collectors of the transistors (Q1, Q2) of the input stage are kept at respective constant voltage levels (VR1, VR2) by amplifier means (A1, A2) with a high input impedance so that the amplifier means (A1, A2) do not substantially absorb current from the collectors.

11. An amplifier according to Claim 9, characterised in that the input stage of the amplifier means (A1, A2) is constituted by MOS transistors (M8,M5).

12. An amplifier according to any one of Claims 9 and 10, characterised in that the amplifier means (A1, A2) operate from reference voltage levels (VR1, VR2) generated by at least one reference voltage (VBG) which is independent of the common supply line (Vcc).

13. An amplifier according to Claims 1 and 10, characterised in that the amplifier means (A2) also control the gate voltage of the second transistor (M10) included in the third branch of the current-mirror circuit.

14. An amplifier according to Claim 1 and Claim 13, taken in combination.

15. An amplifier according to Claim 1, characterised in that at least one (Q1) of the transistors of the input stage is associated with biasing means (IP4) which have high output impedance and are preferably formed by a transistor, the emitter of which is degenerated by means of a resistance.

16. An amplifier according to any one of the preceding claims, characterised in that it is associated with a probe such as, for example, a lambda probe (10) associated with a catalytic converter (K), the output signal (VS) of the probe being applied between the input terminals (IN(-), IN(+)), and the voltage (VM) of the earth (G') of the probe being generally shifted relative to that of the earth (G) of the amplifier (1).

Ansprüche

1. Verstärker, im besonderen ein Verstärker, der als integrierter Schaltkreis aufgebaut ist, in Kombination mit:

- einer Eingangsstufe, die zwei Transistoren (Q1, Q2) aufweist, die mit konstanten Strömen (IP3, IP4) vorgespannt sind, wobei ihre jeweiligen Kollektoren auf konstanten, im allgemeinen positiven, festen Spannungspegeln (VR1, VR2) gehalten werden, so daß ihre jeweiligen Emitter, die die Eingangsanschlüsse (IN(+), IN(-)) des Verstärkers bilden, auf eine Spannung unterhalb des Massepotentials fallen können, wodurch verhindert wird, daß die Substratdiode leitet, wobei die Basiselektroden der beiden Transistoren (Q1, Q2) über einen ersten Widerstand (R1) verbunden sind, so daß im Betrieb eine Spannung, die dem Signal (VIN) entspricht, das zwischen die Eingangsanschlüsse (IN(+), IN(-)) gelegt wird, an den Anschlüssen des ersten Widerstands (R1) liegt,

- einer Stromspiegelschaltung (Q5, Q6, Q7, M7, M9 und M10), die eine erste Einrichtung (M7) aufweist, um im ersten Widerstand (R1) einen ersten Strom (IR1) zu erzeugen, der jener Spannung entspricht, die an die Anschlüsse des ersten Widerstands (R1) gelegt wird, sowie einen zweiten Stromgenerator (M9) aufweist, der mit dem ersten Stromgenerator (M7) in einer Spiegelschaltung verbunden ist, um zumindest einen zweiten Widerstand (R6) mit einem zweiten Strom (IR6) anzuspeisen, der eine Nachbildung des ersten Stroms (IR1) ist, so daß ein Ausgangssignal (VOUT), das dem Eingangssignal (VIN) entspricht, von den Anschlüssen des zweiten Widerstands (R6) abgegriffen werden kann, und

- einer gemeinsamen Versorgungsspannung (Vcc) für die Eingangsstufe (Q1, Q2) und für die Stromspiegelschaltung (Q5, Q6, Q7, M7, M9 und M10),

   dadurch gekennzeichnet, daß der Verstärker aufweist:

- eine Verstärkerstufe (A2), mit der der erste Stromgenerator (M7) in Abhängigkeit vom Eingangssignal (VIN) gesteuert wird, und daß

- die Stromspiegelschaltung einen ersten, zweiten und dritten Schaltkreiszweig aufweist, wobei:

- der erste Schaltkreiszweig (M7, Q5, R3) den ersten Stromgenerator (M7) bildet und den Ausgang des Stromspiegels darstellt, der den ersten Strom (IR1) im ersten Widerstand (R1) erzeugt,

- der zweite Schaltkreiszweig (M9, Q7, R5) den zweiten Stromgenerator (M9) bildet und mit dem ersten Schaltkreiszweig (M7, Q5, R3) als Stromspiegelschaltung in einem im allgemeinen "kaskodenartigen" Aufbau verbunden ist, und

- der dritte Schaltkreiszweig (M10, Q6, R4) im Aufbau zum ersten Schaltkreiszweig (M7, Q5, R3) und zum zweiten Schaltkreiszweig (M9, Q7, R5) symmetrisch ist, wobei der dritte Schaltkreiszweig zwischen der gemeinsamen Versorgungsspannung (Vcc) und der Masse (G) des Schaltkreises liegt, wobei der erste Schaltkreiszweig (M7, Q5, R3) zwischen der gemeinsamen Versorgungsspannung (Vcc) und dem ersten Widerstand (R1) liegt, und wobei der zweite Schaltkreiszweig (M9, Q7, R5) zwischen der gemeinsamen Versorgungsspannung (Vcc) und dem zweiten Widerstand (R6) liegt, und wobei weiters:

jeder Schaltkreiszweig der Strompspiegelschaltung:

- ein erster Transistor (Q5, Q6, Q7) mit seinem Emitter an der gemeinsamen Versorgungsspannung (Vcc) liegt, während seine Basis mit der Basis eines homologen Transistors der anderen Schaltkreiszweige der Stromspiegelschaltung verbunden ist, und

- ein zweiter Transistor (M7, M10, M9), der vorzugsweise von einem MOS-Transistor gebildet wird, mit seiner Quelle am ersten Transistor (Q5, Q6, Q7) und mit seiner Steuerelektrode an den Steuerelektroden der homologen Transistoren der Schaltkreiszweige der anderen Stromspiegelschaltung liegt,

   wobei die Basis des ersten Transistors (Q6), der im dritten Schaltkreiszweig (M10, Q6, R4) liegt, mit einem anderen Anschluß des ersten Transistors (Q6) zu einem im allgemeinen diodenartigen Aufbau kurzgeschlossen ist,
inkludiert.

2. Verstärker gemäß Anspruch 1, dadurch gekennzeichnet, daß ein Anschluß des zweiten Widerstands (R6) mit Masse verbunden ist.

3. Verstärker gemäß Anspruch 1 oder 2, dadurch gekennzeichnet, daß den beiden Transistoren der Eingangsstufe (Q1, Q2) Konstantstrom-Vorspannungsstufen (IP3, IP4) zugeordnet sind, die konstante Vorspannungsströme im µA-Bereich liefern.

4. Verstärker gemäß Anspruch 3, dadurch gekennzeichnet, daß den beiden Transistoren der Eingangsstufe (Q1, Q2) Konstantstrom-Vorspannungsstufen (IP3, IP4) zugeordnet sind, die einen konstanten Vorspannungsstrom unter 1 µA, vorzugsweise in der Größenordnung von 0,5 µA liefern.

5. Verstärker gemäß irgendeinem der bisherigen Ansprüche, dadurch gekennzeichnet, daß die beiden Transistoren in der Eingangsstufe (Q1, Q2) bipolare Transistoren sind.

6. Verstärker gemäß Anspruch 1, dadurch gekennzeichnet, daß die Senken der zweiten Transistoren (M7, M9), die im ersten (M7, Q5, R3) und zweiten (M9, Q7, R5) Schaltkreiszweig der Stromspiegelschaltung enthalten sind, mit dem ersten Widerstand (R1) bzw. mit dem zweiten Widerstand (R6) verbunden sind.

7. Verstärker gemäß Anspruch 1, dadurch gekennzeichnet, daß die Senke des zweiten Transistors (M10), der im dritten Schaltkreiszweig (M10, Q6, R4) der Stromspiegelschaltung enthalten ist, mit der Masse (G) der Stromspiegelschaltung verbunden ist.

8. Verstärker gemäß Anspruch 1, dadurch gekennzeichnet, daß der erste Transistor (Q5, Q6, Q7) ein bipolarer Transistor ist.

9. Verstärker gemäß Anspruch 1, dadurch gekennzeichnet, daß der erste Stromgenerator im wesentlichen von einem MOS-Transistor (M7) oder einer anderen ähnlichen Einrichtung mit einem vernachlässigbaren Eingangs-Vorspannungsstrom gebildet wird, um den Diskriminationsfaktor (common-mode rejection ratio) der Eingangsanschlüsse des Verstärkers (IN(+), IN(-)) zu erhöhen.

10. Verstärker gemäß Anspruch 1, dadurch gekennzeichnet, daß die Kollektoren der Transistoren (Q1, Q2) der Eingangsstufe von Verstärkerstufen (A1, A2) mit einer hohen Eingangsimpedanz auf entsprechenden konstanten Spannungspegeln (VR1, VR2) gehalten werden, so daß die Verstärkerstufen (A1, A2) im wesentlichen keinen Strom von den Kollektoren aufnehmen.

11. Verstärker gemäß Anspruch 9, dadurch gekennzeichnet, daß die Eingangsstufe der Verstärkerstufen (A1, A2) von MOS-Transistoren (M8, M5) gebildet wird.

12. Verstärker gemäß irgendeinem der Ansprüche 9 und 10 dadurch gekennzeichnet, daß die Verstärkerstufen (A1, A2) von Bezugsspannungspegeln (VR1, VR2) arbeiten, die von zumindest einer Bezugsspannung (VBG) erzeugt werden, die von der gemeinsamen Spannungsversorgung (Vcc) unabhängig ist.

13. Verstärker gemäß Anspruch 1 und 10, dadurch gekennzeichnet, daß die Verstärkerstufe (A2) weiters die Steuerelektrodenspannung des zweiten Transistors (M10) steuert, der im dritten Schaltkreiszweig der Stromspiegelschaltung enthalten ist.

14. Verstärker gemäß Anspruch 1 und 13 als Kombination.

15. Verstärker gemäß Anspruch 1, dadurch gekennzeichnet, daß zumindest einem (Q1) der Transistoren der Eingangsstufe eine Vorspannungsstufe (IP4) zugeordnet ist, die eine hohe Ausgangsimpedanz besitzt und vorzugsweise von einem Transistor gebildet wird, dessen Emitter mit Hilfe eines Widerstands entartet ist.

16. Verstärker gemäß irgendeinem der bisherigen Ansprüche, dadurch gekennzeichnet, daß dem Verstärker eine Sonde, beispielsweise eine Lambda-Sonde (10), zugeordnet ist, die einem katalytischen Umsetzer (K) zugeordnet ist, wobei das Ausgangssignal (VS) der Sonde zwischen die Eingangsanschlüsse (IN(+), IN(-)) gelegt wird, und wobei die Spannung (VM) der Masse (G') der Sonde im allgemeinen relativ zur Masse (G) des Verstärkers (1) versetzt ist.

Revendications

1. Amplificateur, en particulier amplificateur intégré, comprenant, en combinaison:

- un étage d'entrée comprenant deux transistors (Q1, Q2) polarisés par des courants constants (IP3, IP4), avec leurs collecteurs respectifs maintenus à des niveaux de tension fixe, généralement positive, constante (VR1, VR2), de sorte que leurs émetteurs respectifs, qui définissent les bornes d'entrée (IN (+), IN(-)) de l'amplificateur, puissent atteindre une tension inférieure à celle de la terre, empêchant la diode de substrat de conduire, les deux transistors (Q1, Q2) étant connectés par leurs bases respectives au moyen d'une première résistance (R1) de sorte que, en fonctionnement, une tension correspondant au signal (VIN) appliqué entre les bornes d'entrée (IN(+), IN(-)) soit appliquée aux bornes de la première résistance (R1),

- un circuit miroir de courant (Q5, Q6, Q7, M7, M9 et M10) comprenant des premiers moyens (M7) destinés à générer, dans la première résistance (R1), un premier courant (IR1) correspondant à la tension appliquée aux bornes de la première résistance (R1) et des seconds moyens de génération de courant (M9) qui sont couplés aux premiers moyens de génération de courant (M7) dans une relation de miroir afin d'alimenter au moins une seconde résistance (R6) avec un second courant (IR6) qui constitue une réplique du premier courant (IR1), de sorte qu'un signal de sortie (VOUT) correspondant au signal d'entrée (VIN) puisse être pris à partir des bornes de la seconde résistance (R6), et

- une ligne d'alimentation commune (Vcc) pour l'étage d'entrée (Q1, Q2) et pour le circuit miroir de courant (Q5, Q6, Q7, M7, M9 et M10), caractérisé en ce qu'il comprend

   des moyens formant amplificateur (A2) qui commandent les premiers moyens de génération de courant (M7) en fonction du signal d'entrée (VIN), et en ce que

- ledit circuit miroir de courant comprend des premier, second et troisième branchements, dans lequel:

- le premier branchement (M7, Q5, R3) définit les premiers moyens de génération de courant (M7) et est la sortie du miroir de courant générant ledit premier courant (IR1) dans ladite première résistance (R1),

- le second branchement (M9, Q7, R5) définit les seconds moyens de génération de courant (M9) et est couplé au premier branchement (M7, Q5, R3) dans une relation de miroir de courant selon une configuration généralement de type "cascade", et

- le troisième branchement (M10, Q6, R4) est de structure symétrique au premier branchement (M7, Q5, R3) et au second branchement (M9, Q7, R5), le troisième branchement étant interposé entre la ligne d'alimentation commune (Vcc) et la terre (G) du circuit, le premier branchement (M7, Q5, R3) étant interposé entre la ligne d'alimentation commune (Vcc) et la première résistance (R1) et le second branchement (M9, Q7, R5) étant interposé entre la ligne d'alimentation commune (Vcc) et la seconde résistance (R6) et, de plus, dans lequel:

   chaque branchement du circuit miroir de courant comprend:

- un premier transistor (Q5, Q6, Q7) avec son émetteur connecté à la ligne d'alimentation commune (Vcc) et sa base connectée aux bases des transistors homologues des autres branchements du circuit miroir de courant, et

- un second transistor (M7, M10, M9), de préférence constitué par un transistor MOS, avec sa source connectée au premier transistor (Q5, Q6, Q7) et sa grille connectée aux grilles des transistors homologues des branchements du circuit miroir de courant,

   la base du premier transistor (Q6) compris dans le troisième branchement (M10, Q6, R4) étant court-circuitée à l'une des autres bornes du premier transistor (Q6) selon une configuration généralement semblable à une diode.

2. Amplificateur selon la revendication 1, caractérisé en ce qu'une borne de la seconde résistance (R6) est connectée à la terre.

3. Amplificateur selon la revendication 1 ou la revendication 2, caractérisé en ce que les deux transistors de l'étage d'entrée (Q1, Q2) sont associés à des circuits de polarisation de courant constant (IP3, IP4) qui délivrent des courants de polarisation constants dans la plage de µA.

4. Amplificateur selon la revendication 3, caractérisé en ce que les deux transistors de l'étage d'entrée (Q1, Q2) sont associés à des circuits de polarisation de courant constant (IP3, IP4) qui délivrent un courant de polarisation constant inférieur à 1 µA, de préférence de l'ordre de 0,5 µA.

5. Amplificateur selon l'une quelconque des revendications précédentes, caractérisé en ce que les deux transistors dans l'étage d'entrée (Q1, Q2) sont des transistors bipolaires.

6. Amplificateur selon la revendication 1, caractérisé en ce que les drains des seconds transistors (M7, M9) compris dans les premier (M7, Q5, R3) et second (M9, Q7, R5) branchements du circuit miroir de courant sont, respectivement, connectés à la première résistance (R1) et à la seconde résistance (R6).

7. Amplificateur selon la revendication 1, caractérisé en ce que le drain du second transistor (M10) compris dans le troisième branchement (M10, Q6, R4) du circuit miroir de courant est connecté à la terre (G) du circuit miroir de courant.

8. Amplificateur selon la revendication 1, caractérisé en ce que le premier transistor (Q5, Q6, Q7) est un transistor bipolaire.

9. Amplificateur selon la revendication 1, caractérisé en ce que les premiers moyens de génération de courant sont constitués essentiellement par un transistor MOS (M7) ou tout autre dispositif similaire avec un courant de polarisation d'entrée négligeable, afin d'augmenter le taux de réjection en mode commun des entrées (IN(-), IN(+)) de l'amplificateur.

10. Amplificateur selon la revendication 1, caractérisé en ce que les collecteurs des transistors (Q1, Q2) de l'étage d'entrée sont maintenus à des niveaux de tension constante respectifs (VR1, VR2) par des moyens formant amplificateur (A1, A2) avec une impédance d'entrée élevée de sorte que les moyens formant amplificateur (A1, A2) n'absorbent pas sensiblement le courant provenant des collecteurs.

11. Amplificateur selon la revendication 9, caractérisé en ce que l'étage d'entrée des moyens formant amplificateur (A1, A2) est constitué des transistors MOS (M8, M5).

12. Amplificateur selon l'une quelconque des revendications 9 et 10, caractérisé en ce que les moyens formant amplificateur (A1, A2) fonctionnent à partir de niveaux de tension de référence (VR1, VR2) générés par au moins une tension de référence (VBG) qui est indépendante de la ligne d'alimentation commune (Vcc).

13. Amplificateur selon les revendications 1 et 10, caractérisé en ce que les moyens formant amplificateur (A2) commandent également la tension de grille du second transistor (M10) compris dans le troisième branchement du circuit miroir de courant.

14. Amplificateur selon la revendication 1 et la revendication 13, combinées.

15. Amplificateur selon la revendication 1, caractérisé en ce qu'au moins un (Q1) des transistors de l'étage d'entrée est associé à des moyens de polarisation (IP4) qui ont une impédance de sortie élevée et qui sont formés, de préférence, par un transistor dont l'émetteur est dégénéré au moyen d'une résistance.

16. Amplificateur selon l'une quelconque des revendications précédentes, caractérisé en ce qu'il est associé à une sonde, telle que, par exemple, une sonde lambda (10) associée à un convertisseur catalytique (K), le signal de sortie (VS) de la sonde étant appliqué entre les bornes d'entrée (IN(-), IN(+)) et la tension (VM) de la terre (G') de la sonde étant généralement décalée par rapport à celle de la terre (G) de l'amplificateur (1).

Drawing